Fast Extremum-Seeking Based Robust State Estimation and Control for Frequency Support of Low-Inertia Microgrids

Inhalt:
Traditional control methods for fast frequency support (FFS) of microgrids require high-fidelity models and offer limitations. Fast extremum-seeking control (FESC) ensures reduced model dependence and has shown potential for such applications. However, the robust performance of FESC is restricted due to existing model-dependent state estimation techniques used in microgrids, especially when the system undergoes dynamic changes due to parameter variations. This paper proposes a fast extremum-seeking based estimation (FESE) technique to complement the robustness of FESC with similar design properties for FFS of microgrids. Simulations on a microgrid test system in Matlab/Simulink show overall lower Normalized Root Mean Square Error (NRMSE) with FESE for tracking frequency-dependent states compared to conventional modellight ensemble Kalman filter (EnKF) during dynamic changes in system parameters. The proposed FESE-FESC scheme ensures system frequency stability with lower steady-state deviation during a frequency event compared to the EnKF-FESC scheme. In addition, the presented FESC in the paper incorporates a penalty-function-based constraint input mechanism to ensure optimization is performed respecting operational limits on control action. The results illustrate the proposed extremum-seeking based estimation and control framework as a robust and modellight alternative for microgrid frequency support.